Most of the time i just use that:<p><a href="https://pgtune.leopard.in.ua/#/" rel="nofollow">https://pgtune.leopard.in.ua/#/</a>
For me, the greatest pain of postgresql is upgrading between major versions. You need to install both versions to the same db server at the same time and then Tthe data must be copied (using pg_upgrade) between the two clusters, something that leads to downtime (that is proportional to the size of the data ie how much time it needs to copy and run some maintenance commands). If you consider that our cluster hosts databases for like 15 apps and each needs to be changed to display some "maintenance mode" home page you'lln understand my frustration and fear of major upgrades...<p>I know there are some ways that reduce the downtime, like hard linking the data between the clusters instead of version, or using logical replication between the clusters and switching to the new version cluster. The problem is that since I am not a full time DBA (and my organisation doesn't have one) I don't trust myself with these techniques and rely to the trusted pg_upgrade method that will leave the old cluster as it was with all its data intact in case anything went wrong!<p>The only good thing is that postgresql supports each major version for a lot of time (5 years) so such updates don't need to be very frequent :)
I use this almost daily. Last time it was an hour ago -- checked when hot_standby's default changed from off to on (it was in Postgres 10).<p>Highly recommend it.
Does anyone know if there's a way to pull secrets into PG from an environment variable or similar and access it within PG in a way that does not directly expose the value to the db user or to logs?
Any guides on how to combine SSD's, software RAID (Linux md) and ext4 for best performance? Plenty of guides for the individual components but not all of them together
Has anyone other than Amazon successfully migrated off oracle.. and ported stuff into Postgres... e.g. 1000s of PL/SQL procedures/packages moved to postgres